Face Recognition Algorithms for Online and On-Site Classes

Abstract

Face recognition is used in a wide variety of applications such as surveillance systems, human-computer interaction, automatic door access control systems, and network security. One of the policies of the smart university is to adopt technology to help with teaching and learning, especially during the Covid-19 pandemic. In this paper, a smart attendance system using face recognition algorithms with deep learning is proposed and used in the university's classroom. Instead of calling names to confirm the identity of students, our system does it automatically. The system was tested in 3 scenarios, namely, in online classes, in on-site classes, and in problematic cases using a standard dataset. The performances of the 3 scenarios were compared in the experiment in terms of precision, recall, F1 score, and percentage accuracy. Our result revealed that in online classes the recognition accuracy is as high as 100%. The implemented system is inexpensive and practical. The application can be used on any portable device such as tablets or smartphones. History viewing, multiple subjects handling, and file exporting features are also incorporated into the system.